Parameter identification algorithm for fault location based on distributed transmission line model

This paper proposed a novel time-domain fault location algorithm using one-terminal data. By using parameter identification method and distributed transmission line model, fault can be located with local voltage and current data. Since the fault distance, fault resistance and the remote terminal source parameters are considered as unknown variables, the time- domain equations of the fault-superimposed component network are solved to determine the fault distance. The technique uses two-dimensional search in the plane of fault distance and fault resistance. At any searching step, the two variables to be searched are assumed to be fixed values. Therefore, a series of remote terminal source parameters are calculated with moving data windows. In the end, both fault distance and fault resistance, with which the divergence from the calculated opposite source parameters is minimum, are regarded as the actual parameters. This algorithm eliminates influence of the fault resistance and the source parameters at the opposite end on the fault location accuracy, which has been the main cause of error for conventional one-terminal approaches. In addition, this algorithm uses the transient sampling data directly with short data window. The EMTP simulation results prove the validity and high accuracy of the new algorithm.